DE102020212298A1 - Method and device for device-specific determination of an internal resistance-based aging state of a battery - Google Patents

Abstract

The invention relates to a method for determining an internal resistance-based state of health (SOHR) of a battery in a device, with the following steps: - Determining the internal resistance-based state of health (SOHR) using a hybrid state of health model (1), the hybrid state of health model (1) being a physical aging model ( 2) and includes a correction model, the physical aging model (2) being based on an electrochemical battery model in order to provide a physical aging state (SOHRph) depending on one or more operating variables (I, T, U, SOC) of the battery, the Correction model (3) is designed based on data in order to map operating characteristics derived from one or more operating variables (I, T, U, SOC) to a correction variable, with the physical aging condition (SOHRph) being subjected to the correction variable (k) in order to internal resistance-based aging state agree, - determining a profile of a modeled battery voltage (Um) using an adaptation model (7), which is modeled depending on internal states of the physical aging model (2) and depending on a profile of at least one battery current (I), - providing the modeled battery voltage (Um) to the physical aging model (2) in order to determine the physical aging state (SOHR) depending on the modeled battery voltage (Um).

Description

technical field

The invention relates to batteries, in particular for electrically driven motor vehicles, and also to measures for determining an aging state based on a rate of change in internal resistance in accordance with a customer, manufacturer or device specification.

Technical background

Electrically drivable motor vehicles are usually supplied with electrical energy via vehicle batteries. The aging status of the vehicle battery decreases over the course of its service life, which results in a decreasing maximum storage capacity. The degree of aging of the vehicle battery depends on the individual load on the vehicle battery.

In the case of batteries, the state of health is the key variable for specifying a remaining battery capacity or remaining battery charge. The state of health represents a measure of the aging of the battery since it was put into service (beginning of life). The state of health can be specified as a capacity retention rate (SOHC) or as an increase in internal resistance (SOHR) according to a rate of change. The capacity retention rate SOHC is given as the ratio of the measured instantaneous capacity to an initial capacity of the fully charged battery.

The relative change in the internal resistance SOHR increases as the battery ages. When determining the internal resistance-dependent aging state, different approaches are provided, depending on battery-specific internal resistance values, for determining the internal resistance-based aging state. Thus, the specification of the internal resistance-based aging state is significantly dependent on the determination rule, which can be specified device-specifically or manufacturer-specifically.

Disclosure of Invention

According to the invention, a method for determining an internal resistance-based aging state of a battery according to claim 1 and a corresponding device according to the independent claim are provided.

Further developments are specified in the dependent claims.

• - Bestimmen eines innenwiderstandsbasierten Alterungszustands mithilfe eines hybriden Alterungszustandsmodells mit einem elektrochemischen Alterungsmodell und einem Korrekturmodell, wobei das elektrochemische Alterungsmodell ausgebildet ist, um einen physikalischen Alterungszustand abhängig von einer oder mehreren Betriebsgrößen der Batterie bereitzustellen, wobei das Korrekturmodell datenbasiert ausgebildet ist, um aus der einen oder den mehreren Betriebsgrößen abgeleitete Betriebsmerkmale auf eine Korrekturgröße abzubilden, wobei der physikalische Alterungszustand mit der Korrekturgröße beaufschlagt wird, um den innenwiderstandsbasierten Alterungszustand zu bestimmen,
• - Ermitteln eines Verlaufs einer modellierten Batteriespannung mithilfe eines Anpassungsmodells, der abhängig von internen Zuständen des physikalischen Alterungsmodells und abhängig von einem Verlauf von zumindest einem Batteriestrom modelliert wird,
• - Bereitstellen der modellierten Batteriespannung an das physikalische Alterungsmodell, um den physikalischen Alterungszustand zu ermitteln.

According to a first aspect, a method for determining an internal resistance-based aging state of a battery is provided, with the following steps:

• - Determining an internal resistance-based aging state using a hybrid aging state model with an electrochemical aging model and a correction model, the electrochemical aging model being designed to provide a physical aging state depending on one or more operating parameters of the battery, the correction model being data-based in order to consist of one or mapping the operating characteristics derived from the several operating variables to a correction variable, the correction variable being applied to the physical aging condition in order to determine the internal resistance-based aging condition,
• - Determining a profile of a modeled battery voltage using an adaptation model that is modeled depending on internal states of the physical aging model and depending on a profile of at least one battery current,
• - Providing the modeled battery voltage to the physical aging model in order to determine the physical aging condition.

The aging state can be determined based on the internal resistances of an electronic equivalent circuit of a battery. However, the internal resistances of the battery are defined differently on a customer-specific basis, so that one difficulty is to consider the internal resistance-based aging state on these different definitions using the same algorithm platform.

The state of health of a battery can be determined using a hybrid state of health model. This envisages using a physical aging model that determines an aging condition based on operating parameters and an electrochemical battery model. The electrochemical battery model describes the battery behavior using a set of differential equations that describe electrochemical equilibrium states and characterizes the battery condition in terms of a variety of internal states.

Furthermore, a data-based correction model is provided, which provides a correction variable depending on operating characteristics. The operating characteristics are from the courses of one or more operating variables of the battery, such as the Course of the battery current, and in particular the course of the battery voltage, the course of the battery temperature and / or the course of the state of charge derived. Furthermore, the operating characteristics can be derived based on internal states of the physical aging model. The correction variable is used to apply the physical state of health in order to obtain the current state of health.

In addition to other variables, a battery is often characterized by its internal resistance, which results from an equivalent circuit assumed for a battery. The determination of these internal resistances is defined in different ways depending on the device manufacturer, so that the specification of the aging condition based on this depends significantly on the definition of the determination of the internal resistances.

The adjustment model can be designed to determine the profile of the modeled battery voltage as a function of the profile of the battery current based on differential equations based on an electronic equivalent circuit.

One idea of the above method is that, in addition to the operating variables, the physical aging model is supplied with a modeled battery voltage (modeled terminal voltage) obtained from an adaptation model, which results from a simulation of the operating variables with the internal resistance values of the battery equivalent circuit, which have been determined according to the device-specific definition.

Provision can be made for the adaptation model to carry out the course of the modeled battery voltage as a function of at least one internal resistance, with the at least one internal resistance being determined from one or more internal states of the physical aging model.

The adaptation model uses a simple electrical battery model based on the internal resistances and the way they are connected in the battery equivalent circuit and simulates a corresponding battery voltage depending on the way the internal resistances have been determined. For example, internal resistances are specified to calculate the aging condition as dU/dl for a specific constant current, for a specified period of e.g. 10s and at a specific battery temperature of e.g. T=25°C. This means that a specific voltage difference is determined for a specific constant current flow for a specified period of time in order to determine the corresponding internal resistance. According to the equivalent circuit model, the way in which the internal resistance is determined therefore represents an influence on the battery voltage, which is evaluated in the state of health model in order to determine the state of health.

Thus, the adaptation model serves to balance different ways of determining the internal resistance of the battery, so that the same architecture of the hybrid state of health model can always be used.

In order to also enable the aging model based on reference data, which indicate a measured aging condition of the battery at a certain time, to take into account the device specification for determining the internal resistance, the electrochemical aging model can be adapted based on the reference data by the internal states of the physical Aging model are adjusted depending on a difference between the modeled aging condition and the measured aging condition according to the reference data.

Thus, when measuring an aging state, at least one of the internal states of the physical aging model can be adjusted depending on a difference between the modeled internal resistance-based aging state and the measured aging state.

Provision can also be made for the data-based correction model to be retrained if sufficient training data is available.

character list

• 1 ein Blockdiagramm zur Veranschaulichung einer Funktion zum Ermitteln eines kundenspezifischen innenwiderstandsbasierten Alterungszustands;
• 2 ein Flussdiagramm zur Veranschaulichung eines Verfahrens zum Betreiben eines Alterungszustandsmodells zum Ermitteln eines kundenspezifischen Alterungszustands; und
• 3 eine Ersatzschaltung der Batterie zur Veranschaulichung der Verschaltung der Innenwiderstände.

Embodiments are explained in more detail below with reference to the accompanying drawings. Show it:

• 1 a block diagram for illustrating a function for determining a customer-specific internal resistance-based aging state;
• 2 a flowchart to illustrate a method for operating a state of health model for determining a customer-specific state of health; and
• 3 an equivalent circuit of the battery to illustrate the wiring of the internal resistances.

Description of Embodiments

1 FIG. 1 shows a schematic block diagram of an aging state model 1 for providing a device-specific internal resistance based aging condition SOHR according to individual calculation specifications of one or more internal resistances.

The aging state model 1 is constructed in a hybrid manner and has a physical aging model 2 at its core, which provides a physical aging state. For this purpose, curves of operating variables of the battery, namely the curves of battery current I, battery voltage U, battery temperature T and state of charge SOC, are used as input variables. The physical aging model 2 is designed as an electrochemical battery model and uses differential equations to describe the corresponding electrochemical states Z1, Z2, Z3..., such as layer thicknesses (e.g. SEI thickness), changes in the cyclable lithium due to anode/cathode side reactions, rapid consumption of Electrolytes, slow consumption of electrolytes, loss of active material in anode, loss of active material in cathode, and the like. The physical aging model 2 indicates a physical aging state SOHRph corresponding to the curves of the operating parameters of the battery and depending on the internal electrochemical states Z1, Z2, Z3.

However, the model values for the physical aging state SOHR provided by the electrochemical model are imprecise in certain situations, and it is therefore intended to correct them with a correction variable k. For this purpose, a data-based correction model 3 is provided, which is designed as a machine learning model, such as a Gaussian process model. The correction model 3 is trained to map operating features M and the physical aging state SOHRph to the correction variable k.

In a correction block 4, the physical aging state SOHRph is corrected with the correction variable k in order to output the aging state SOHR based on the internal resistance. The correction quantity k can be applied additively, multiplicatively or in some other way.

Operating characteristics M are derived from the operating variables I, U, T, SOC and from the internal electrochemical states Z1, Z2, Z3 . . . of the aging model 2 in an operating characteristics block.

The operating characteristics M relate to an evaluation period. The evaluation period for determining the aging state can be a few hours (e.g. 6 hours) to several weeks (e.g. one month). A usual value for the evaluation period is one week.

The operating characteristics can include, for example, characteristics relating to the evaluation period and/or accumulated characteristics and/or statistical values ascertained over the entire service life to date. In particular, the operating characteristics can include, for example: electrochemical states (layer thicknesses, concentrations, cyclable lithium, ...), histogram data over the course of the state of charge, the temperature, the battery voltage, the battery current, in particular histogram data regarding the battery temperature distribution over the state of charge, the distribution of the charging current over the temperature and/or the discharge current distribution over temperature, accumulated total charge (Ah), an average increase in capacity during a charge (especially for charges where the increase in charge is above a threshold proportion (e.g. 20%) of the total battery capacity), a maximum of differential capacity (dQ/dU: change in charge divided by change in battery voltage) and others.

Further information can be derived from the operating characteristics: a load pattern over time such as charging and driving cycles, determined by usage patterns (such as fast charging at high currents or strong acceleration or braking processes with recuperation), a usage time of the vehicle battery, a charge accumulated over the running time and a discharge accumulated over the term, a maximum charging current, a maximum discharging current, a charging frequency, an average charging current, an average discharging current, a power throughput during charging and discharging, a (especially average) charging temperature, a (especially average) spread of the state of charge and the like.

An internal resistance determination block 6 is also provided, which detects the internal states Z1, Z2, Z3 from the physical aging model and calculates them according to a suitable calculation rule R1=f(Z1, Z2, Z3...) and R2=g(Z1, Z2, Z3 ...) into the internal resistances R1 and R2, which characterize the battery in its electronic equivalent circuit. The conversion functions g(), f() can be provided as a parameterized polynomial function, as a data-based function model or the like. The conversion is based on the assumption that the relationship between the resistances R1, R2 and the states Z1, Z2, Z3 is linear. The parameters that describe the linear relationship can be fitted to the training data.

The physical aging model 2 is designed to obtain curves of operating parameters of the battery, in particular the battery current Battery voltage, the battery temperature and the current state of charge. In order to also use the structure of the aging state model for other types of calculations of the aging state SOHR, an adaptation model 7 is now provided to model a battery voltage Um that takes into account the changed calculation of the internal resistance-based aging state.

For this purpose, the internal resistances R1, R2 determined in the internal resistance block 6 are fed to the adaptation model 7, which contains an electrical battery model based on an electronic equivalent circuit of the battery. The battery model uses the structure of the electronic equivalent circuit of the battery to model the course of the battery voltage Um, depending on the course of the operating variables I, U, T, SOC and the current state of aging SOHR, which the measured battery voltage U (terminal voltage) from the operating variables replaced. Such a battery replacement circuit is in 3 shown schematically with an ideal voltage source V, the internal resistances R1, R2 and a capacitance C. Further RC elements can be provided. As a result, the physical aging model 2 is operated in such a way that a physical aging state SOHRph adapted according to the changed device specification is determined.

This makes it possible to use the hybrid aging state model 1 for different requirements for determining the aging state with regard to the internal resistance-based aging state SOHR.

If reference data for the aging status are obtained from battery operation, these are used to adjust the internal statuses Z1, Z2, Z3 . . . of the physical aging model. The reference data correspond to measured aging states that are used to adapt the physical aging model 2 as soon as possible after their measurement.

wobei SOHR_mess dem gemessenen Alterungszustand aus den Referenzdaten entspricht. ε₁, ε₂, ε₃ entsprechen einer Gewichtung und Lernrate, mit der der interne Zustand Z1, Z2, Z3... des physikalischen Alterungsmodells 2 angepasst wird. Dies führt über den Innenwiderstandsblock 6 dazu, dass auch die Innenwiderstände R1, R2 in dem Anpassungsmodell 7 verändert werden, um die Berechnung der modellierten Batteriespannung Um an die veränderten Innenwiderstände R1, R2 anzupassen. Insgesamt sind die Innenwiderstände R1 und R2 das Ergebnis des elektrochemischen Alterungsmodells, ergänzt mit der datenbasierten Modellkorrektur.The adaptation of the physical aging model 2 takes place accordingly in steps $$\begin{array}{l}Z\end{array}$$$$\begin{array}{l}1\leftarrow \mathrm{<figure-callout\; id="1"\; label="Z"\; filenames="DE102020212298A1\_0002.png,DE102020212298A1\_0003.png"\; state="\{\{state\}\}">Z</figure-callout>}1+{\mathrm{<figure-callout\; id="1"\; label="e"\; filenames="DE102020212298A1\_0002.png,DE102020212298A1\_0003.png"\; state="\{\{state\}\}">e</figure-callout>}}_1\cdot \left({\text{SOHR SOHR}}_{\text{mess}}\right)\\ \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="DE102020212298A1\_0002.png"\; state="\{\{state\}\}">Z</figure-callout>}2\leftarrow \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="DE102020212298A1\_0002.png"\; state="\{\{state\}\}">Z</figure-callout>}2+{\mathrm{<figure-callout\; id="2"\; label="e"\; filenames="DE102020212298A1\_0002.png"\; state="\{\{state\}\}">e</figure-callout>}}_2\cdot \left({\text{SOHR SOHR}}_{\text{mess}}\right)\\ \mathrm{<figure-callout\; id="3"\; label="Z"\; filenames="DE102020212298A1\_0002.png"\; state="\{\{state\}\}">Z</figure-callout>}3\leftarrow \mathrm{<figure-callout\; id="3"\; label="Z"\; filenames="DE102020212298A1\_0002.png"\; state="\{\{state\}\}">Z</figure-callout>}3+{\mathrm{<figure-callout\; id="3"\; label="e"\; filenames="DE102020212298A1\_0002.png"\; state="\{\{state\}\}">e</figure-callout>}}_{{}_3}\cdot \left({\text{SOHR SOHR}}_{\text{mess}}\right)\\ ...\end{array}$$

where SOHR _mess corresponds to the measured aging condition from the reference data. ε ₁ , ε ₂ , ε ₃ correspond to a weighting and learning rate with which the internal state Z1, Z2, Z3 . . . of the physical aging model 2 is adjusted. Via the internal resistance block 6, this means that the internal resistances R1, R2 are also changed in the adaptation model 7 in order to adapt the calculation of the modeled battery voltage Um to the changed internal resistances R1, R2. Overall, the internal resistances R1 and R2 are the result of the electrochemical aging model, supplemented with the data-based model correction.

The method for operating the state of health model 1 is described below with reference to the flow chart 2 described in more detail. The method is executed in a battery control device and is implemented there in hardware and/or software.

In step S1, the operating variables I, T, SOC of the battery are recorded, namely the battery current, the battery temperature and the current state of charge.

The operating variables are used in the adaptation model 7 in step S2 in order to determine the modeled battery voltage Um. This is based on specified internal resistances R1, R2 and using differential equations of a known electrical battery model that can map the battery current I, the state of charge SOC and the battery temperature T to a battery voltage Um.

The operating variables I, T, SOC, the modeled battery voltage Um and the current aging state SOHR are supplied to the physical aging model 2 in step S3. The physical aging model 2 uses electrochemical differential equations to determine the equilibrium states of the battery for modeling a physical aging state SOHRph.

Furthermore, in step S4 , operating characteristics are generated from the operating variables I, U, T, SOC in the operating characteristics block 5 and are supplied to the correction model 3 . The internal states Z1, Z2, Z3 of the physical aging model 2, which are also used to determine operating characteristics, are also supplied to the operating characteristics block 5.

From the operating characteristics and the physical aging state SOHRph, a correction variable k is determined in the correction model 3 in step S5, which is applied to the physical aging state SOHRph in the correction block 4 in order to obtain the current aging state SOHR.

The current state of health SOHR is provided to the adaptation model in step S6 in order to continuously determine the battery voltage according to the current state of health.

In step S7, the internal resistances R1 and R2, which are used in the adaptation model 7 for modeling the battery voltage Um, are derived from the internal states Z1, Z2, Z3, . . .

In step S8 it is checked whether reference data are present. The reference data include a state of health SOH _mess measured in the battery. If the measured state of health SOH _mess deviates from the current state of health SOHR by more than a predetermined threshold value (alternative: yes), the method continues with step S9, otherwise (alternative: no) the process jumps back to step S1.

In step S9, the internal states Z1, Z2, Z3 of the physical aging model are adjusted, in particular in accordance with the above specification, and the method is then continued with step S10.

In step S10, a check is made, e.g. using a threshold value comparison with the number of training data sets, to determine whether sufficient training data is available for retraining the correction model. If this is the case (alternative: yes), the method continues with step S10, otherwise the process jumps back to step S1.

In step S11, the correction model 3 is updated with the new training data or trained further, the training data comprising a series of training data sets for different evaluation periods. The training data sets correspond to assignments of measured aging states to operating feature points as a combination of operating features and internal states of the physical aging model 2. A jump is then made back to step S1.

Claims (10)

Method for determining an internal resistance-based state of health (SOHR) of a battery in a device, with the following steps: - Determining the internal resistance-based state of health (SOHR) using a hybrid state of health model (1), the hybrid state of health model (1) a physical aging model (2) and a Correction model includes, wherein the physical aging model (2) is designed based on an electrochemical battery model to provide a physical aging state (SOHRph) depending on one or more operating variables (I, T, U, SOC) of the battery, wherein the correction model (3) is designed based on data in order to map operating characteristics derived from one or more operating variables (I, T, U, SOC) onto a correction variable, the physical aging condition (SOHRph) being applied with the correction variable (k) in order to determine the internal resistance-based aging condition , - Identify a s course of a modeled battery voltage (U _m ) using an adjustment model (7), which is modeled depending on internal states of the physical aging model (2) and depending on a course of at least one battery current (I), - providing the modeled battery voltage (U _m ) to the physical aging model (2) in order to determine the physical aging state (SOHR) depending on the modeled battery voltage (U _m ). procedure after claim 1 , wherein the adaptation model (7) is designed to determine the course of the modeled battery voltage (U _m ) as a function of the course of the battery current based on differential equations based on an electronic equivalent circuit. procedure after claim 1 or 2 , wherein the adjustment model (7) makes the profile of the modeled battery voltage (U _m ) dependent on at least one internal resistance (R1, R2), wherein the at least one internal resistance (R1, R2) depends on one or more internal states of the physical aging model (2) is determined. Procedure according to one of Claims 1 until 3 , wherein when measuring an aging state, at least one of the internal states of the physical aging model (2) is adjusted depending on a difference between the modeled internal resistance-based aging state (SOHR) and the measured aging state. Procedure according to one of Claims 1 until 4 , the data-based correction model (3) being trained if sufficient training data is available. Procedure according to one of Claims 1 until 5 , The method being carried out in a central unit to which the device provides the progression of the operating variables. Procedure according to one of Claims 1 until 6 , wherein the device with the specific battery corresponds to a motor vehicle, a pedelec, an aircraft, in particular a drone, a machine tool, an IoT device, an autonomous robot and/or a household appliance. Apparatus for determining an internal resistance-based aging state of a battery, comprising the following steps: - determining the internal resistance-based aging state (SOHR) using a hybrid aging state model (1), the hybrid aging state model (1) comprising a physical aging model (2) and a correction model, the physical Aging model (2) is designed based on an electrochemical battery model in order to provide a physical state of aging (SOHRph) depending on one or more operating variables (I, T, U, SOC) of the battery, the correction model (3) being designed based on data in order to mapping the operating characteristics derived from one or more operating variables (I, T, U, SOC) to a correction variable, the physical aging condition (SOHRph) being applied with the correction variable (k) in order to determine the internal resistance-based aging condition, - determining a course of a model ellated battery voltage (U _m ) using an adaptation model (7), which is modeled depending on internal states of the physical aging model (2) and depending on a course of at least one battery current (I), - providing the modeled battery voltage (U _m ) to the physical aging model (2) to determine the physical aging condition (SOHR) depending on the modeled battery voltage (U _m ). Computer program product comprising instructions which, when the program is executed, cause at least one data processing device to carry out the steps of the method according to one of Claims 1 until 7 to execute. Machine-readable storage medium, comprising instructions which, when executed by at least one data processing device, cause the latter to carry out the steps of the method according to one of Claims 1 until 7 to execute.

Images